摘要
振动模态分析技术和神经网络技术是解决结构损伤诊断问题的有效工具。分别对组合结构、框架结构的动力参数进行计算,并模拟组合结构的各种单损伤情况、框架结构的单损伤和双损伤情况的各种组合,提出了修正隐含层层数公式和优选振型向量的方法。说明了用低阶振型和经优选后的振型向量结合的方法进行组合结构、框架结构损伤诊断的可行性。
Vibration modal analysis technology and neural network technology is a effective instrument to solve damage diagnose.It was numerically simulated all kinds of one-damage identification and two damage identification in a combined structure and a frame structure.It was proposed the method of how to choose the best number of implication layers and how to optimize mode of vibration vector.Which showed the fassility of using low order frequency and certain mode of vibration or their derived quantity incorporation by way of structural impair position and extent of damage.
出处
《工业建筑》
CSCD
北大核心
2006年第z1期228-230,共3页
Industrial Construction
关键词
损伤识别
振型
频率
神经网络
对称结构
隐含层
优选振型向量
damage identification mode of vibration frequency neural network symmetrical structure implication layer optimize mode of vibration vector